1AUBIOONSET(1)                 aubio User's manual                AUBIOONSET(1)
2
3
4

NAME

6       aubioonset - a command line tool to extract musical onset times
7

SYNOPSIS

9       aubioonset source
10       aubioonset [[-i] source] [-o sink]
11                  [-r rate] [-B win] [-H hop]
12                  [-O method] [-t thres]
13                  [-T time-format]
14                  [-s sil] [-m] [-f]
15                  [-j] [-N miditap-note] [-V miditap-velo]
16                  [-v] [-h]
17
18

DESCRIPTION

20       aubioonset  attempts  to  detect onset times, the beginning of discrete
21       sound events, in audio signals.
22
23       When started with an input  source  (-i/--input),  the  detected  onset
24       times are given on the console, in seconds.
25
26       When  started  without  an  input  source,  or  with  the  jack  option
27       (-j/--jack), aubioonset starts in jack mode.
28

OPTIONS

30       This program follows the usual GNU command line syntax, with  long  op‐
31       tions  starting  with two dashes (--). A summary of options is included
32       below.
33
34       -i, --input source
35              Run analysis on this audio  file.  Most  uncompressed  and  com‐
36              pressed are supported, depending on how aubio was built.
37
38       -o, --output sink
39              Save results in this file. The file will be created on the model
40              of the input file. Onset times are marked by a short  wood-block
41              like sound.
42
43       -r, --samplerate rate
44              Fetch  the  input  source, resampled at the given sampling rate.
45              The rate should be specified in Hertz as an integer. If  0,  the
46              sampling  rate  of the original source will be used. Defaults to
47              0.
48
49       -B, --bufsize win
50              The size of the buffer to analyze, that is  the  length  of  the
51              window  used for spectral and temporal computations. Defaults to
52              512.
53
54       -H, --hopsize hop
55              The number of samples between  two  consecutive  analysis.   De‐
56              faults to 256.
57
58       -O, --onset method
59              The  onset detection method to use. See ONSET METHODS below. De‐
60              faults to 'default'.
61
62       -t, --onset-threshold thres
63              Set the threshold value for the onset peak picking.  Values  are
64              typically  in  the  range [0.001, 0.900]. Lower threshold values
65              imply more onsets detected. Increasing this threshold should re‐
66              duce the number of incorrect detections. Defaults to 0.3.
67
68       -M, --minioi value
69              Set  the  minimum inter-onset interval, in seconds, the shortest
70              interval between two consecutive onsets. Defaults to 0.020
71
72       -s, --silence sil
73              Set the silence threshold, in dB, under which the onset will not
74              be  detected.  A  value of -20.0 would eliminate most onsets but
75              the loudest ones. A value of -90.0 would select all onsets.  De‐
76              faults to -90.0.
77
78       -T, --timeformat format
79              Set time format (samples, ms, seconds). Defaults to seconds.
80
81       -m, --mix-input
82              Mix source signal to the output signal before writing to sink.
83
84       -f, --force-overwrite
85              Overwrite output file if it already exists.
86
87       -j, --jack
88              Use  Jack  input/output.  You  will  need a Jack connection con‐
89              troller to feed aubio some signal and listen to its output.
90
91       -N, --miditap-note
92              Override note value for MIDI tap. Defaults to 69.
93
94       -V, --miditap-velop
95              Override velocity value for MIDI tap. Defaults to 65.
96
97       -h, --help
98              Print a short help message and exit.
99
100       -v, --verbose
101              Be verbose.
102

ONSET METHODS

104       Available methods are:
105
106       default
107              Default distance, currently hfc
108
109       Default: 'default' (currently set to hfc)
110
111       energy Energy based distance
112
113       This function calculates the local energy of the input spectral frame.
114
115       hfc    High-Frequency content
116
117       This method computes the High Frequency  Content  (HFC)  of  the  input
118       spectral  frame.  The resulting function is efficient at detecting per‐
119       cussive onsets.
120
121       Paul Masri. Computer modeling of Sound for Transformation and Synthesis
122       of Musical Signal. PhD dissertation, University of Bristol, UK, 1996.
123
124       complex
125              Complex domain onset detection function
126
127       This function uses information both in frequency and in phase to deter‐
128       mine changes in the spectral content that might correspond  to  musical
129       onsets.   It  is  best  suited  for  complex signals such as polyphonic
130       recordings.
131
132       Christopher Duxbury, Mike E. Davies, and Mark B. Sandler.
133              Complex domain onset detection for musical signals. In  Proceed‐
134              ings  of  the  Digital  Audio Effects Conference, DAFx-03, pages
135              90-93, London, UK, 2003.
136
137       phase  Phase based onset detection function
138
139       This function uses information both in frequency and in phase to deter‐
140       mine  changes  in the spectral content that might correspond to musical
141       onsets. It is best  suited  for  complex  signals  such  as  polyphonic
142       recordings.
143
144       Juan-Pablo Bello, Mike P. Davies, and Mark B. Sandler.
145              Phase-based  note onset detection for music signals. In Proceed‐
146              ings of the IEEE International Conference  on  Acoustics  Speech
147              and Signal Processing, pages 441444, Hong-Kong, 2003.
148
149       specdiff
150              Spectral difference onset detection function
151
152       Jonhatan  Foote and Shingo Uchihashi. The beat spectrum: a new approach
153       to rhythm analysis. In IEEE International Conference on Multimedia  and
154       Expo (ICME 2001), pages 881884, Tokyo, Japan, August 2001.
155
156       kl     Kulback-Liebler onset detection function
157
158       Stephen  Hainsworth  and Malcom Macleod. Onset detection in music audio
159       signals. In Proceedings of the International Computer Music  Conference
160       (ICMC), Singapore, 2003.
161
162       mkl    Modified Kulback-Liebler onset detection function
163
164       Paul  Brossier, ``Automatic annotation of musical audio for interactive
165       systems'', Chapter 2, Temporal segmentation,  PhD  thesis,  Centre  for
166       Digital music, Queen Mary University of London, London, UK, 2006.
167
168       specflux
169              Spectral flux
170
171       Simon Dixon, Onset Detection Revisited, in ``Proceedings of the 9th In‐
172       ternational Conference on Digital Audio Effects'' (DAFx-06),  Montreal,
173       Canada, 2006.
174

SEE ALSO

176       aubiopitch(1),  aubiotrack(1),  aubionotes(1),  aubioquiet(1),  aubiom‐
177       fcc(1), and aubiocut(1).
178

AUTHOR

180       This manual page was written by Paul Brossier <piem@aubio.org>. Permis‐
181       sion  is  granted to copy, distribute and/or modify this document under
182       the terms of the GNU General Public License as published  by  the  Free
183       Software  Foundation,  either version 3 of the License, or (at your op‐
184       tion) any later version.
185
186
187
188aubio 0.4.9                      20 July 2023                    AUBIOONSET(1)
Impressum